EpiVax discovers and develops epitqpe-driven vaccines by screening microbial genomes for protective epitopes. In this application, we describe a genome-based approach for the development of a therapeutic vaccine against Helicobacter pylori disease. First, the H. pylori genome will be scanned for immunodominant Th epitopes by computer-driven (EpiMatrix) analysis. Candidate Th epitopes will be selected based on class II MHC allele restriction and clustering (promiscuity). Epitopes derived from antigens previously shown in vivo as protective (e.g., enzymes, porins, adhesins, heat shock proteins), as well as novel epitopes from proteins not previously implicated in protection will be further sorted on the basis of strain conservation (> 95%), in vivo expression in human and murine gastric tissue, and display of minimal (<30%) sequence homology with human proteins. Peripheral blood leukocytes from H. pylori-infected subjects, and spleen cells from H. pylori-immunized HLA transgenic mice will be used to examine and confirm the ability of the epitopes to activate Th1 and/or Th2 immune responses in vitro. Selected epitopes will be aligned and reverse translated in a string-of-beads array into a DNA construct. The epitope construct will then be cloned into DNA plasmids optimized for expression in vivo. The vaccine candidates will be systematically examined for their ability to confer therapeutic protection from infection in HLA transgenic mice, and for their ability to avert the development of gastric cancer in a p27-/- mutant mouse model. Results from these studies will identify novel vaccines for control of H. pylori infection, and for the first time, examine directly the protective effect of vaccination against H. pylori-associated malignant disease. Phase I is devoted to feasibility; a Phase II SBIR program would examine the best means of enhancing immunogenicity, using reduction of chronic infection and modulation of metaplastic disease in well-established mouse models of H. pylori disease as our correlate of efficacy.